VFD for Centrifugal Chiller

[TECMSC]

Hello All,

We are installing two new centrifugal chillers (1000T and 500T) to an existing chiller plant (2400t). Existing chillers or pumps do not have VFD's. System is Primary pumping-no primary secondary loop. During this addition we are considering adding VFD's for the new chillers.

Reliability is paramount concern although we are looking for ways to save energy as well.

I need to develop a list of pros vs cons for adding VFD's to the new chillers. Any ideas would be greatly appreciated.

My initial take is:
Pro -
1. Save energy at low load for first year (although in a year or 1 1/2 we should be at full load and not need VFD)
2. Save energy during economizer cycle since we would be reducing the load on the chiller plant i.e. be at part load on chiller plant but full load in space.
3. ??

Cons
1. VFD less reliable than other starting methods (eg reduced voltage starter)
2. VFD or VFD parts availability is questionable. VFD fails and chiller goes down; finding a VFD/VFD-part could take weeks.
3. Reliable integration of VFD w/Chiller historically has been poor.
4. Mean time to failure for VFD is shorter than other starting methods.
5 Potentially more maintenance.
6. ??

One thought is to provide both a VFD and reduced voltage starter in tandem with a xxx pole xxx throw switch to transfer power through either the VFD or alternate starter. If the VFD fails we would always be able to provide power through the alternate starter. Problem is that I'm not sure if the chiller manufacturer would warranty this type of starting option.

Comments/Thoughts greatly appreciated??

 

[ChrisConley]

Just a quick comment, a VFD doesn't save energy on low load, it saves energy on low lift.

You could also consider adding VFD's to your primary pumps to save energy, and while you're at it, VFD's to the cooling tower.

Throw in a Hartford loop for controling the multiple VFD's.

 

[TECMSC]

Thanks for the reply.

I thought low lift was the same as low load. I'll have to look further to fully understand this concept.

The cooling tower fans have VFDs. The primary pumps might get VFDs because the system could be changed to a varible primary pumping arrangement.

Hartford loop-the existing chillers will not get VFDs, only the new ones. When I did a search on the hartford loop I got back info on steam boilers! Thsi dosen't mean I looked in the right spot!!

 

[ChrisConley]

Mistyped that sorry, Hartman Loop:

http://www.hartmanco.com/innovate/loop/index.htm

Didn't mean to be obscure, the Hartman Loop is a proprietary control strategy that optimizes energy performance of chiller plants with multiple VFD's.

Low lift is the difference between condenser water temperature and evaporator water temperature. Low lift would occur if you have a cooler, or dryer day and your condenser water is less than 90F.

 

[sterl]

York has been using liquid cooled drives on centrifugals for more than 15- years and their experiences on a variety of applications would add a lot to your list. Liquid cooled VFD's are pretty reliable, and will shave some mechanical maintenance out of the compressor and motor bearing requirements if this plant has a diverse or disproportionate capacity or duty cycle.

Bypass contactors are a pretty ordinary part of the kit where the heat yield from the liquid cooling is not useful in any other way. For a fully VFD plant, bypass arrangements introduce some interesting variables, especially if the VFD's will overdrive the synch speed. And if you are going to set up based on bypass contactor, the system's control needs to recognize the bypass and function in spite of it.

The biggest maintenance issue with VFD plants is the motors, largely because the manufacturers have applied their own upgrades to bearings, lubrication arrangements and re-lube cycles, insulation systems, cooling systems and similar. Getting parts for an ABB or similar Liq Cooled VFD is not that bad; knowing where to source an identical or better motor quickly, or getting a shop to rebuild the motor to original specifications, will be a considerable challenge.

 

[TECMSC]

ChrisConley
No problem...I need ask: Lift is the tempreature difference between the evaporator entering temp (CHWR e.g 54degF) and the condenser water entering temp (CWS e.g. 85degF).

Thanks for the Hartman info.

Sterl
The load is 24/7 and there not plans to retro fit the existing chillers with VFDs, only the new ones so a diverse duty cycle isn't a concern -- unless going into economizer is considered diverse duty cycle. I could see a potential motor problem if we retro fitted the existing chillers with VFDs.


The alternate starting method would be manual...just need to know if the VFD caused a chiller shut down so we can restart the chiller through the alternate starter.
Thanks for your input.

-----------------not sure if i should start another thread, but since this is all realted...-----------------

What would be some concerns with changing a primary CHW system to a variable primary CHW system??

Chillers have a min and max flow rate requirement. My concern is the existing chillers(no VFDs) "somemhow" get out of the min/max flow range. What would be a prefered way to regulate flow through existing chillers? (VFDs on CHW pumps to control delta pressure??...I can't see installing a flow regulating valve to the inlet of a chiller)

 

[Dymalica]

I thought you could get VFDs with a bypass, that will basically allow the items to run at full speed in event of a VFD failure.